This is a revised proposal for the project, High Risk Latino Men: An Empowerment Program to Reduce Unsafe Sex. While new AIDS case raters are declining among Anglo gay men in New York City, they are increasing dramatically among Latino men who have se with men (LMSM). Our prior three studies with LMSM have described many traumatic experiences in the lives of LMSM, including heavy stigmatization, harassment, discrimination, violence, and even childhood sexual abuse, which may result in feelings of disempowerment and low critical consciousness, the desire for immediate gratification, and the presence of self-justifications that interfere with consistent safer-sex behavior. We have developed a prevention program based on our qualitative and quantitative studies and organized under the theoretical framework of Freire's Pedagogy of the Oppressed. Our empowerment program consists of eight sessions that focus on critical experiences likely to have occurred in the lives of LMSM, eliciting personal associations, promoting critical analysis about the consequences that such experiences had in the life of the participants, and discussing their feelings of disempowerment. This is followed by a period of collective problem solving to find alternative, more empowering ways to deal with everyday demands, including safer sex behavior. The process is facilitated, but not prescribed, by an activist or education. We propose to test this prevention program in a randomized two-arm trial (experimental versus control). Specifically, we will evaluate the efficacy of an empowerment intervention for defining personal goals; enhancing self-efficacy and intentionality; decreasing substance use, pleasure orientation, attribution of risk to circumstantial events, and beliefs in attenuating factors; and, through these mediators, decreasing unprotected anal sex. Two hundred adult LMSM who have had at least one episode of unprotected anal sex in he past year will be randomly assigned to either the experimental or the control group. Primary analyses, utilizing randomization tests will assess the intervention effectiveness at reducing unprotected anal sex. Subsequent analyses will explore the role of mediating and background variables using generalized estimating equations.